pH response and mechanical properties of Fe2O3–TeO2-based glass/stainless steel enamel electrodes for pH sensors

Glass pH sensors are unsuitable for in vivo biomedical, clinical, or food applications because of the brittleness of glass and the difficulty in measuring small volumes. Enamel structures such as glass/stainless steel are candidates for glass-based pH electrodes. In this study, new enamel electrodes for pH sensors using Fe2O3–TeO2-based glass/stainless steel were developed. The effect of NiO addition to Fe2O3–TeO2 glass on the pH sensitivity and the three-point bending strength of enamels were investigated. The effect of NiO addition to Fe2O3–TeO2 glass/stainless steel on the pH sensitivity was negligible. Fe2O3–TeO2-based glass/stainless steel showed pH sensitivity appropriate to a working electrode. Enameling at a lower temperature under an air atmosphere was desirable for narrowing the gap between pH 4–7 and pH 7–9 sensitivities. The NiO addition to Fe2O3–TeO2 glass/stainless steel decreased the three-point bending strength. Therefore, NiO did not serve as an adhesion oxide in the Fe2O3–TeO2 glass. Fe2O3–TeO2 glass/stainless steel possessed the highest three-point bending strength among all samples when prepared at 670 °C under an air atmosphere. Therefore, no NiO addition and enameling at a lower temperature under an air atmosphere are desirable for obtaining more robust Fe2O3–TeO2 glass/stainless steel than Li2O–SiO2-based glass electrodes for pH sensors.


Introduction
Commercially available pH combination electrodes often consist of the following electrodes: (1) a working electrode that generates an electromotive force in response to the concentration of hydrogen ions in the solution and (2) a reference electrode. We have developed several working electrodes [1][2][3][4][5][6][7]. Recently, Ag and Ag alloy-precipitated Ag 2 O-TeO 2 glass and Ag 2 O-TeO 2 glass/stainless steel enamel reference electrodes for pH sensors were developed [8].
The NiO and CoO are used as adhesion oxides between glass and stainless steel in typical enamels [44][45][46][47][48]. It has not been revealed whether NiO serves as an adhesion oxide in Fe 2 O 3 -NiO-TeO 2 glass or not. Thus, Fe 2 O 3 -TeO 2 glass-based glasses were selected as new glass/stainless steel enamels in the present study. Furthermore, the effect of NiO addition to Fe 2 O 3 -TeO 2 enamels on the pH sensitivity and the three-point bending strength were investigated.

Preparation of 20FexNi(80-x)Te glass/stainless steel enamels
The 20FexNi(80-x)Te glasses were crushed and classified as 20FexNi(80-x)Te glass powder of less than 53 μm. The 20FexNi(80-x) Te glass powder with a thickness of 1.0 mm was deposited on SUS304 stainless steel (Nilaco Corporation, Tokyo, Japan) with a thickness of 0.2 mm using the dry-type doctor blade method. Then, 20FexNi(80-x)Te glass/stainless steel enamels (abbreviated 20FexNi(80-x)Te/SUS) were obtained via heat-treatment at 670-730 • C for 0.5 h at the programming rate of 100 • C/min with (N 2 ) and without (air) an N 2 flow rate of 0.5 L/min, and then furnace cooling to fuse the 20FexNi(80-x)Te glass powder onto the stainless steel. The stainless steel is a back electrode that supports the 20FexNi(80-x)Te glasses at high heat-treatment temperatures because the stainless steel does not contact the test solution in potentiometric measurements.

Three-point bending strength measurement of samples
The mechanical properties of 20FexNi(80-x)Te glasses and 20FexNi(80-x)Te/SUS were measured using a ZTA-500 N digital force gauge (Imada Co., Ltd., Toyohashi, Japan). The BT-500 N three-point bending test fixture was set to a distance between the fulcrums of 10 mm (Imada Co., Ltd., Toyohashi, Japan). The MX2-1000-FA standard vertical motorized test stand was set to a cross-head speed of 2.5 mm/min (Imada Co., Ltd., Toyohashi, Japan) to evaluate the three-point bending strength.

Thermal properties of 20FexNi(80-x)Te glasses
The results of the DTA measurement of 20FexNi(80-x)Te glasses are summarized in Table 1. The glass transition temperature (T g ), crystallization peak temperature (T p ), and T p -T g related to the thermal stability of the glasses increased via NiO addition, whereas the melting temperature (T m ) decreased. The T g and T p measured under an air atmosphere were higher than those measured under an N 2 atmosphere. The transformation of Fe 2+ into Fe 3+ occurred under an air atmosphere. As a result, the viscosity of the glass increased, and crystallization became difficult [49][50][51]. This change in the valence of Fe ions under an air atmosphere is consistent with our XRD data, where the precipitation of Fe 2 O 3 is suppressed, as will be seen in Section 3.2.2. Fig. 2 presents the changes in potential with the measurement time in pH 7, pH 4, and pH 9 buffer solutions for 20FexNi(80-x)Te glass/SUS prepared at 730 • C. The change in potential with the measurement time corresponding to a pH change is significant for all samples, suggesting that it serves as a working electrode. The effect of the NiO addition and atmosphere for preparing glass/stainless steel enamels on pH response seems negligibly small. Fig. 3 shows the relationship between potential and pH (pH 4, pH 7, and pH 9) for 20Fe80Te glass/SUS (670 • C, air) as an example determined by the potential curve. The solid line is average data for three cycles, and the broken line is data for the third cycle. Good linear relationships with high coefficients of determination were observed for both lines. pH sensitivity of 94.0% was obtained from the relationship for 20Fe80Te glass/SUS (670 • C, air).

pH response of 20FexNi(80-x)Te Glass/SUS
The pH responsivity (pH sensitivity, pH repeatability, and pH response time) of 20FexNi(80-x)Te glasses and 20FexNi(80-x)Te glass/SUS are listed in Table 2. Under "pH sensitivity", the column entitled pH 4-9 presents the pH sensitivity between pH 4 and 9. The pH 4-7 sensitivity was higher than the pH 7-9 sensitivity for all samples. 20FexNi(80-x)Te glasses showed lower pH 7-9 sensitivity than 20FexNi(80-x)Te glass/SUS, but a difference occurred between pH 4-7 sensitivity and pH 7-9 sensitivity. The pH 4-9 sensitivity for 20Fe80Te glass/SUS (670 • C, air) in Table 2 was slightly different from that in Fig. 3 because two pHs determine the former, and three pHs determine the latter. The appropriate pH measurement of 20FexNi(80-x)Te glasses and 20FexNi(80-x)Te glass/SUS was conducted [8] because the electrical resistivity is less than 10 10 Ω･cm, and a high pH sensitivity was obtained. The effect of the NiO addition on pH sensitivity was negligible for both 20FexNi(80-x)Te glasses and 20FexNi(80-x)Te glass/SUS. Enameling at a lower temperature under an air atmosphere was desirable for narrowing the gap between pH 4-7 and pH 7-9 sensitivities.

Three-point bending strength of samples
The corrected load-displacement curves for determining the three-point bending strength of 20FexNi(80-x)Te glass/SUS prepared at 730 • C are given in Fig. 4. In this figure, the curve of stainless steel heat-treated under the same atmosphere was subtracted from that Table 1 Glass transition temperature (T g ), crystallization peak temperature (T p ), T p -T g , and melting temperature (T m ) of 20FexNi(80-x)Te glasses. of 20FexNi(80-x)Te/SUS. The maximum load in these curves was defined as the three-point bending strength. Glass/stainless steel enamels heat-treated under air showed higher three-point bending strength than that heat-treated under N 2 . Table 3 lists the three-point bending strength of 20FexNi(80-x)Te/SUS and related samples. The NiO addition to Fe 2 O 3 -TeO 2 glass/ SUS decreased the three-point bending strength. Therefore, NiO did not serve as an adhesion oxide in the Fe 2 O 3 -NiO-TeO 2 glass system. In contrast, the glass/SUS prepared under an air atmosphere showed higher three-point bending strength than those prepared under an N 2 atmosphere. The effect was significant in 20Fe80Te/SUS (air). In addition, 20Fe80Te/SUS (air) possessed the highest three-point bending strength among all samples when prepared at 670 • C. The load-displacement curve for bare stainless steel did not show yield points and similar curves despite the atmosphere used. Therefore, the upper crystallized glass layers dominate the threepoint bending strength of the glass/SUS. The effect of the precipitated crystals on the three-point bending strength is discussed in the following section.        The three-point bending strength of 20Fe80Te/SUS (N 2 ) was one-third of that of 20Fe80Te/SUS (air). The peak intensity of α-Fe 2 O 3 on the surface of 20Fe80Te/SUS (N 2 ) was much higher than that of 20Fe80Te/SUS (air). It is well known that a decrease in Fe 3+ decreases glass viscosity [49][50][51]. Therefore, glass fusion under a N 2 atmosphere induces glass crystallization. Our results are consistent with previous reports. Therefore, α-Fe 2 O 3 may be the starting point of destruction and decrease the three-point bending strength. A significant difference in the crystallized glass was not seen in the present case. The crystallinity of FeCr 0⋅29 Ni 0⋅16 C 0.06 due to SUS304 was high for 20Fe80Te/SUS (air). However, the negative effect of FeCr 0⋅29 Ni 0⋅16 C 0.06 on the three-point bending strength seems negligible. The three-point bending strength of 20Fe80Te/SUS (730 • C, air) was half that of 20Fe80Te/SUS (670 • C, air). The peak intensity of β-TeO 2 on the surface 20Fe80Te/SUS (730 • C, air) was higher than that of 20Fe80Te/SUS (670 • C, air). Therefore, β-TeO 2 may be the starting point of destruction and decrease the three-point bending strength. A significant difference in the crystallized glass was not seen in the present case. The crystallinity of FeCr 0⋅29 Ni 0⋅16 C 0.06 due to the SUS304 of 20Fe80Te/SUS (730 • C, air) was stronger than that of 20Fe80Te/SUS (670 • C, air). However, the effect of the enhanced crystallinity of stainless steel at the crystallized glass and stainless steel interface on the three-point bending strength seems negligible, as in the case of NiO addition (Fig. 4) and atmosphere control (Fig. 5).

Effect of enamel preparation temperature on three-point bending strength
On the other hand, an enameling temperature (630 • C) that was too low resulted in a low three-point bending strength because of low glass fusion. According to these results, no NiO addition and enameling at appropriate low enameling temperatures under an air atmosphere are desirable for obtaining robust enamels. These conditions suppressed the formation of α-Fe 2 O 3 and α-TeO 2 on the surface of enamels. Furthermore, 20Fe80Te/SUS (670 • C, air) was much more robust than commercially available Li 2 O-SiO 2 -based glass electrodes for pH sensors. Our result may show that Fe 2 O 3 -containing materials, such as novel ferrite nanoparticles [52], are candidates for new pH and other electrodes.

Conclusions
New enamel electrodes for pH sensors using Fe 2 O 3 -TeO 2 -based glass/SUS were developed in this study. The effect of NiO addition to Fe 2 O 3 -TeO 2 glass/SUS on the pH sensitivity and the three-point bending strength were investigated.
• The effect of NiO addition to Fe 2 O 3 -TeO 2 glass/SUS on pH sensitivity was negligible. Fe 2 O 3 -TeO 2 -based glass/SUS showed appropriate pH sensitivity as a working electrode. Enameling at a lower temperature under an air atmosphere was desirable for narrowing the gap between pH 4-7 and pH 7-9 sensitivities.   • The NiO addition to Fe 2 O 3 -TeO 2 glass/SUS decreased the three-point bending strength. Therefore, NiO did not serve as an adhesion oxide in the Fe 2 O 3 -TeO 2 glass. In contrast, Fe 2 O 3 -TeO 2 glass/SUS possessed the highest three-point bending strength among all samples when prepared at 670 • C under an air atmosphere. These conditions suppressed the formation of α-Fe 2 O 3 and α-TeO 2 on the surface of enamels. • No NiO addition or enameling at a lower temperature under an air atmosphere is desirable for obtaining robust enamels.
20Fe80Te/SUS (670 • C, air) was much more robust than commercially available Li 2 O-SiO 2 -based glass electrodes for pH sensors.

Funding statement
This work was supported by JSPS KAKENHI (JP18K04702 and JP22K04685).

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments
The authors thank Miss Yuumi Aoki for her helpful work.